the relationship between language and thought in motion event...
TRANSCRIPT
https://doi.org/10.30961/lr.2019.55.3.457 457
The Relationship between Language and Thought in Motion Event Cognition in Second Language AcquisitionSoo-Ok Kweon†
POSTECH
ABSTRACTThis study aims to investigate the relationship between language and thought in motion events encoding (linguistic perspective) and the perception (non-linguistic perspective) of Korean learners of English. The purpose is to determine whether L2 learners will shift toward L2-based conceptual representations when they can switch the linguistic behavior of encoding motion from their L1 (Korean) to L2 (English). A verbal description task to test linguistic switch and a similarity rating task to test conceptual change are used with animations that describe motion events. Results show that although the linguistic switch to the L2 canonical type of description occurs, the switch does not result in cognitive restructuring by L2 learners when language is not involved. The results will be discussed with respect to “thinking for speaking” and the linguistic relativity dubbed the Sapir-Whorf hypothesis.
Keywords: motion events, L2 acquisition, thinking for speaking, linguistic relativity, language and thought
1. Introduction
One of the persistent questions in second language (L2) acquisition is that if L2
learners can speak as fluently in L2 as in L1, would it be possible for them to make
conceptual shifts in how they think about events in the world; i.e., does learning
an additional language in adulthood affect the L2 learners’ conceptual representation?
This question of language and conceptualization examines whether language can
shape the way people think even if they do not think for speaking. The culmination
of this question on the interaction of language and thought was the Sapir-Whorf
Hypothesis ([1940] 2012) based on linguistic determinism. The strong version of the
* This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2017S1A5A2A01026703).
† Corresponding author: [email protected]
Copyright ⓒ 2019 Language Education Institute, Seoul National University. This is an Open Access article under CC BY-NC License (http://creative-commons.org/licenses/by-nc/4.0).
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon458
Sapir-Whorf Hypothesis holds that “differences among languages cause differences
in the thoughts of their speakers” (p. 65). The linguistic deterministic idea of this
hypothesis is that language can determine the way in which people think, even if
they do not use language in the thoughts. This strong version of Sapir-Whorf
Hypothesis implies implicitly that people think in language-specific ways, even when
they are not actively using language during non-linguistic cognitive processing
(Brown & Lenneberg,1954; Lucy, 1992). This version of the Sapir-Whorf Hypothesis
is overly deterministic and is not supported by evidence (Swoyer, 2011).
Another view is what we will call the weak version of Sapir-Whorf Hypothesis.
Thinking for speaking (Slobin, 1996; 2000; 2004) account is a modified weak version
of the Sapir-Whorf Hypothesis. Thinking for speaking is based on linguistic
relativity, which emphasizes the influence of language on conceptual representation
only during the processing of speaking (i.e., ‘intimately tied to language’, Slobin,
1996, p. 75). Thinking for speaking does not concern the influence of language on
non-linguistic cognition in general (see Bassetti & Cook, 2011 for discussion). Slobin’s
proposal is different from the linguistic determinism (i.e., strong Sapir-Whorf
Hypothesis) claim in that Slobin specifies how thought is to be shaped by language-
specific patterns during online language processing, whereas the linguistic determinism
proposal allows thoughts to be invariably influenced by language, whether one is
speaking or not (Lai, Rodriguez, & Narasimhan, 2014).
Many studies have examined the relationship between language and
conceptualization focusing on motion events mainly, first, in L1 acquisition research
done in linguistic condition observing children learning their native language (Allen,
Ozyrek, Kita, Brown, Furman, Ishizuka & Fujii, 2007; Choi & Bowerman, 1991;
Gentner, 2003; Goldin-Meadow & Zheng, 1998; Maguire & Dove, 2008). Second,
L1 psycholinguistic research involving verbal productions of L1 speakers (Allen et
al., 2007; Athanasopoulos & Bylund, 2013; Berman & Slobin, 1994; Bylund, 2009,
Bylund & Athanasopoulos, 2014; Choi & Bowerman,1991; Oh, 2003; Slobin, 1996;
2000, among others). Third, L2 psycholinguistic research involving verbal production
by L2 learners (Brown & Gullberg, 2013; Cadierno & Ruiz, 2006; Choi & Lantolf,
2008; Hendriks & Hickmann, 2011; Hohenstein, 2006; Lai et al., 2014, among many).
To widen the understanding of the influence of language on thought, the behavior
patterns of language users must be investigated not only in linguistic condition, but
in non-linguistic condition that does not involve use of language. However, this type
of study has been relatively scarce in L1 research (cf. Papafragou, Hulbert, &
Trueswell, 2008; Trueswell & Papafragou, 2010) and almost absent in L2 research
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 459
(cf. Czechowska & Ewert, 2011). None of the published research to date has tested
the switch to L2 in speaking (linguistic) and conceptual restructuring (non-linguistic)
together by adult L2 learners, focusing on motion events.
This study attempts to fill the gap by investigating the relationship between
language and conceptualization in the setting of L2 acquisition and processing of
motion events, from both a linguistic perspective (verbal description) and from a
non-linguistic perspective (perception). More specifically, the idea is to determine,
if linguistic switch occurs in encoding motion events by Korean adult learners of
L2 English, whether the linguistic switch can cause them to also experience
conceptual changes in perceiving the motion events accordingly. The possibility of
conceptual changes in L2 cannot be examined by looking only at language
production in the absence of non-linguistic performance.
2. Theoretical Background
2.1. Motion events
Human languages vary in how they encode descriptions of motion events. The
way people perceive motion events is influenced by the languages they know and
use. Languages can be classified into two broad categories that use different ways
to encode lexicalization patterns of semantic components of a motion event (Talmy,
1985; 2000). A motion event is defined as a change of location, so the key information
about the motion event is its PATH (Talmy, 1985). According to Talmy’s (1985;
2000) typology of languages, languages can be divided into two groups that differ
in how they encode the PATH of motion, i.e., trajectory or the directionality of
the moving object. Satellite-framed languages, including English, German and
Chinese, represent PATH by satellites to the verb such as particles or prepositional
phrases (to the house) and convey MANNER as a main verb (ran) as in (1).
(1) English: He ran into the house.
Verb-framed languages, including Spanish, Turkish and Korean, indicate PATH by
the main verb in a clause: verbs such as entar ‘enter’, salir ‘exit’, and MANNER
of motion are expressed in independent constituents or as adverbial phrases (running)
as in (2).
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon460
(2) Spanish: Entró a la casa corriendo.
‘He entered the house running’
The preferred pattern in such languages is to use the main verb to encode either
“simple motion (e.g., ‘go’) or directed motion (e.g., ‘enter’), then to express
MANNER in an optional adjunct phrase, in those instances, when MANNER is
at issue.” (Slobin, 2000, p. 110).
Korean is a Verb-framed language; it uses verbs to expresses most PATH
information, and uses adverbs to express MANNER separately. In sentence (3), the
main verb in motion clause is a serial verb that is composed of the verb tul ‘enter’
and a deitic verb (i.e., kata ‘go’). This serial main verb indicates motion PATH,
i.e., the core schema of a motion event. The MANNER of motion is encoded by
the optional verb ttwui ‘run.
(3) namca-ka dongkul-lo (ttwui-e) tul-e ka-ss-ta.
Man-SUBJ cave-LOC (run-CONN) enter-CONN go-PST-DECL
[FIGURE] [GROUND] ([MANNER]) [PATH] [DEIXIS]
‘A man went in(to) the cave (running).’
Satellite-framed languages differ systematically from Verb-framed languages in how
they express motion and the concept related to it. Languages such as English have
verbs that conflate motion and MANNER, whereas languages such as Korean have
verbs that conflate motion and PATH. Therefore, Satellite-framed languages (e.g.,
English) are MANNER-oriented languages, whereas Verb-framed languages (e.g.,
Korean) are PATH-oriented languages.
2.2. Research on linguistic description of motion in L1
Numerous studies have examined how Talmy’s (1985, 2000) typological framework
and Slobin’s (1996) thinking for speaking hypothesis affect the expression of motion
by native speakers of various languages. At the most general level, native speakers
of different languages have shown different motion event categorization patterns with
language-specific lexical pressure on event components (i.e., MANNER encoding or
PATH encoding of motion events). Thinking for speaking while comprehending
motion events has been investigated with speakers of MANNER-oriented and
PATH-oriented languages in different verbal production tasks in which they were
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 461
asked to describe events depicted as static pictures, animations or videos. Crucially,
speakers of MANNER-oriented languages (i.e., English) preferred MANNER verbs,
whereas speakers of PATH-oriented languages (i.e., Spanish) preferred PATH verbs.
For instance, Papafragou and Selimis (2010) examined English- and Greek-speaking
adults and five-year-old children and found that English speakers preferred to use
MANNER verbs to describe the video clips of motion events, whereas Greek
speakers preferred PATH verbs. These results showed cross-linguistic differences
between English-and Greek speakers. A comparison of young children who were
acquiring English and Korean (Choi & Bowerman, 1991) determined that children’s
construction of early spatial semantic categories is language-specific, and suggested
that children are guided in the lexicalization patterns of motion events by the
language to which they are exposed from the beginning.
Talmy (2000) assured that PATH is considered to be the “core-schema” of a
motion event. In child language acquisition, some studies proposed universal verb
construal, which holds that children initially show similar conceptualization patterns
regardless of their native language. As knowledge of their native languages increases,
the universal set of event components for the verb system becomes language specific.
Children’s default verb interpretation for a novel verb is to encode the PATH with
significant preference over the MANNER of motion (Allen et al., 2007; Gentner,
2003; Maguire & Dove, 2008). Several naturalistic observations also reported that
children’s utterances include more PATH verbs or PATH components than
MANNER verbs or MANNER components (Allen et al, 2007; Choi & Bowerman,
1991; Goldin-Meadow & Zheng, 1998). Interestingly, however, research findings
incompatible with PATH as “core-schema” of a motion event were observed. Oh
(2003) examined L1 English children (mean age 3:8) and adults and L1 Korean
children (mean age 3:65) and adults to elicit verbal descriptions of video clips
containing motion events. Oh’s (2003) results were contradictory to previous
literature in that both groups of young children produced a high tendency to describe
motion events in terms of MANNER only and to ignore PATH. Even though
children in both groups expressed MANNER of motion more than PATH in their
description, the elaboration of MANNER verbs were limited mostly to walk and
run verbs. Adults, on the contrary, expressed MANNER motion with further
elaboration in half of descriptions by Korean-speaking adults and even more
descriptions by English-speaking adults.
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon462
2.3. Research on linguistic description of motion in L2
Research on L2 learners’ or bilinguals’ production of motion events has shown
diverse results about the possibility of switch to canonical L2 type of behavior and
of L1 retention in thinking for speaking.
The thinking for speaking proposal implies that thought can bring about language-
specific patterns during online processing only when language is involved, but not
when language (i.e., speaking) is not involved. In this respect, Lai et al. (2014) in
a forced similarity judgment test found that the classification of motion events by
Spanish-English bilinguals was affected by the language (either Spanish or English)
that was used to verbally encode the motion event. Hohenstein (2006) examined
early and late L1 Spanish-L2 English bilinguals who watched dynamic videos of motion
events and were asked to describe the action. The bilinguals used more MANNER
verbs in English and PATH verbs in Spanish, which confirms the language-specific
lexicalization patterns between English and Spanish in motion encoding.
Cross-linguistic differences have been observed in different L1s, whereas L1
retention in thinking for speaking has been observed in L2. For instance, in a verbal
description task, English and French monolinguals clearly show cross-linguistic
differences that are typical typological patterns expected from MANNER-oriented
language (English) and PATH-oriented language (French); however, English
speakers of L2 French showed a trace of thinking for speaking pattern of English
(L1) by providing a MANNER verb separately from a PATH verb (Hendriks &
Hickmann, 2011). The interesting interpretation about L1 retention is that L2 leaners
escape from the L2 mode to L1 mode to solve problems in organizing information
in L2 in order to increase the efficiency of communication. Learners will eventually
switch off the L1 mode of thinking for speaking to become entirely native-like.
Using a new mode of behavior, Choi and Lantolf (2008) also investigated the
L1 retention effect in thinking for speaking in gesture. Using an animated cartoon
story by advanced L2 speakers of Korean (L1 English) and L2 speakers of English
(L1 Korean), Choi and Lantolf (2008) found that the typological difference between
English and Korean prohibits L2 speakers from shifting toward L2 thinking for
speaking patterns, even though their level of proficiency in L2 was highly advanced.
L1 thinking for speaking patterns were retained by these advanced L2 learners.
However, the influence of L1 thinking for speaking may not be present in
advanced L2 learners. For instance, Cadierno and Ruiz (2006) examined how
MANNER and PATH of motion could be expressed depending on whether L1/L2
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 463
are MANNER/PATH or PATH/PATH. When asked to write a narration to
describe the ‘frog story’ (Slobin, 1996; 2000), a group of Danish (MANNER-oriented
language) learners of Spanish (PATH-oriented) did not produce MANNER of
motion to a greater extent than did a group of Italian (PATH-oriented language)
learners of Spanish. Moreover, Danish learner group expressed more PATH of
motion than the Italian learner group and the Spanish control group. These results
“seem to support the view of a rather limited role of the L1 thinking for speaking
patterns in advanced adult language learners (p. 207).” The authors hypothesized
that the influence of L1 thinking for speaking might be stronger during the initial
and intermediate stages of acquisition, but gradually disappears as the proficiency
of the learner increases.
In contrast, L2 can influence L1 in descriptions of motion events. For instance,
Brown and Gullberg (2013) found that L2 affected L1 syntactic clausal packaging
of MANNER and PATH by Japanese learners of L2 English. English uses a
MANNER main verb plus a PATH satellite in a single clause, whereas Japanese
uses multiple main verbs, e.g., MANNER main verbs plus PATH main verbs, and
can therefore compose multiple clauses. Aveledo and Athanasopoulos (2016) also
found L2 influence on L1 motion event encoding by early bilingual L1 Spanish
children (aged 5;00-9;00) learning L2 English who used more MANNER verbs and
fewer PATH verbs in their L1 as a result of the influence of L2.
2.4. Research on non-linguistic perception of motion
Considering cross-linguistic differences between MANNER-oriented languages
and PATH-oriented languages in encoding motion events, a reasonable question is
whether the non-linguistic behavior of speakers from these language groups differs
in the same way as the linguistic patterns of their languages. Here, we examine the
interplay between lexicalization of motion and non-linguistic cognition during
thinking without speaking.
In the recognition task (a task in which participants are asked to decide whether
they saw a particular stimulus during the previous experiment), Malt, Sloman and
Gennari (2003), Papafragou, Massey and Gleitman (2002) and Loucks and Pederson
(2011) found no language effects, even when the classification of events was preceded
by verbalization of the stimuli. For instance, Malt et al. (2003) asked L1 Spanish
and L1 English speakers to watch short films portraying motion events. Participants
were asked to point to two out of three pictures of actions that seemed most similar
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon464
to each other. No effect of language-specific lexicalization strategy on visual memory
recognition was found in either the English or Spanish speakers, which could suggest
that linguistic and non-linguistic behaviors are not dependent on each other in a
strong sense of the linguistic relativity hypothesis.
Papafragou and colleagues examined cross-linguistic differences in attending to
dynamic motion event using eye-tracking methodology. Papafragou et al. (2008) and
Trueswell and Papafragou (2010) found that English and Greek adult native speakers
displayed different patterns of visual attention, preferring periphery encoding strategy
in each language: English native speakers spent more time inspecting the PATH
end point rather than the MANNER of motion, whereas Greek native speakers
displayed toward the opposite pattern. Trueswell and Papafragou (2010) called this
tendency reverse-Whorfian effect, to which we will return in discussion.
Papafragou and Selimis (2010) tested the possible effect of language on event
categorization by examining English-and Greek-speaking adults and five-year-old
children. In a similarity judgment task, language specific effects on perceiving the
motion events were observed when the experiment prompt included linguistic
instructions (e.g., Look the X is doing something!). However, when the prompt was
given without linguistic intrusion (e.g., Look!), similar categorization of events
occurred in non-linguistic conditions regardless of the participants’ native language.
They concluded that there are no cross-linguistic differences in the situations that
involve no language use and that the conceptual representation is universal.
To investigate cross-linguistic differences in motion event perception in bilingual
speakers, Czechowska and Ewert (2011) examined whether the non-linguistic behavior
of speakers from English and Polish differs in line with the linguistic lexicalization
patterns of their languages. Although both English and Polish belong to the same
group of Satellite-framed languages that encode MANNER of motion more saliently
than its PATH, English and Polish have crucial differences in motion lexicalization
in that speakers of English will pay more attention to PATH of motion in
comparison to speakers of Polish. Czechowska and Ewert (2011) asked L1 Polish
L2 English bilinguals to rate the similarity between pictures and found that bilinguals
switched toward L2-based conceptual representation, paying more attention to
PATH of motion, which was consistent with English monolinguals’ behavior.
Czechowska and Ewert (2011) also found that the non-linguistic perception was
affected by L2 proficiency, quality and quantity of the exposure to L2.
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 465
3. The Present Study
In the present study, L1 Korean L2 English learners and L1 English speakers
were asked to rate the similarity of motion events and to describe motion events
in short animations that were displayed on a computer monitor. The experiment
was divided into three blocks (Figure 1). In the first block (0L), no verbal
description was required: participants were asked to view two motion events and
to rate the similarity between them. In the second (1L) block, participants were
required to produce verbal descriptions of the events in their L1. In the third block
(2L), all speakers were required to describe the motions in English; i.e., either as
L1 or as L2.
Figure 1. Schematic representation of animation stimulus for verbal description in
1L/2L block.
3.1. Research questions
This experiment aimed to provide data that would bear on the following research
questions and predictions.
1) Do L1 English and Korean speakers differ in the verbal descriptions of motion
events in the linguistic 1L block?
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon466
2) Do L1 Korean L2 English learners switch to the MANNER-oriented description
type that is canonical for English in the linguistic 2L block?
3) If linguistic switch occurs in the 2L block, do L1 Korean L2 English learners
experience cognitive restructuring towards L2-based conceptual representation
of motion events by showing similar conceptualization patterns to L1 English
speakers in the non-linguistic 0L block?
3.2. Research hypotheses
3.2.1. Hypothesis 1 (1L Block)
At the most general level, because of the cross-linguistic difference between
English and Korean lexicalization patterns of motion events, I predict that L1
English and Korean speakers would differ in their verbal description behavior in
the linguistic 1L block: Korean speakers should overwhelmingly favor PATH-oriented
descriptions, whereas English speakers should generally produce MANNER-oriented
descriptions. This prediction is consistent with thinking for speaking (Slobin, 1996;
2004), which argues that language may influence conceptualization of experience
when this experience is linguistically mediated. In the 1L block linguistic condition,
speakers think about the reality in order to encode it for speech.
3.2.2. Hypothesis 2 (2L Block)
When L1 Korean L2 English learners are required to switch to their L2 (English)
in the 2L block, they might either simply switch to the MANNER-oriented description
type that is canonical for English, or they might transfer their native Korean
PATH-oriented descriptions into English or exhibit some variety of mixed behavior.
3.2.3. Hypothesis 3 (0L Block)
The non-linguistic condition, in which language use is not involved in event
categorization, is crucial to the question of the present study: whether conceptual
restructuring can occur in non-linguistic condition when the linguistic switch
occurred in L2 description of motion events. Two possibilities can be predicted. One
possibility is that cognitive restructuring would occur: Korean speakers will shift
towards L2-based conceptual representation, being more sensitive to the SAME
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 467
MANNER alternate in the non-linguistic 0L block.
The other possibility is that cognitive restructuring may fail to occur. Although
linguistic switch to L2 English canonical description type occurred in the linguistic
2L block, Korean speakers will still be more sensitive to the SAME PATH alternate
that is L1-type conceptual representation in the non-linguistic 0L block.
4. Method
4.1. Participants
The participants in this study were 20 Korean learners of English (3 females, 17
males) who were undergraduate students (mean age 25, range: 20-24) and 13 English
native speakers who were either instructors or exchange students (mean age 35,
ranged: 21-45) at a Korean university. None of the Korean learners of English had
lived in English-speaking countries for more than three months. Based on the criteria
of an intramural speaking and writing test and curriculum system developed by the
university, participants in this study were categorized into advanced group.
4.2. Materials
Test items were 16 sets of animations; each set consisted of three related
animations. The total number of animations that participants saw was 48. Each set
consisted of an Original event and two variant events (Figure 2). The Original events
involved entities (persons, animals, or objects) moving along a PATH (e.g., to, from,
under) in a certain MANNER (e.g., the leaf spinning away from a tree).
All test stimuli involved change-of-state events; each of the variants depicted a
specific change to the Original event (Figure 2a). For example, in the SAME MANNER
variant (Figure 2b), the PATH was changed, but the MANNER remained the same
(the leaf spinning down from a tree). In the SAME PATH variant (Figure 2c), the
MANNER of movement was changed, but PATH was kept the same (the leaf blown
away from a tree). Half of the animation sets involved animate agents, and the other
half involved inanimate agents. There were 2 sets of practice animations that did
not involve pure motion events in which the moving character presented spatial
relations. Also, 12 filler items that were designed for another study involving
animations were used.
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon468
(a) Original (b) SAME MANNER (c) SAME PATE
Figure 2. Sample stimuli. From left to right: Original event (leaf spinning away from
a tree), SAME MANNER variant (leaf spinning down from a tree), SAME PATH
variant (leaf blown away from a tree).
4.3. Procedure
Participants were asked to watch the Original animation presented on a computer
screen, remembering the animation, then press a button when they were ready to
see a new animation, which caused the Original baseline animation to disappear
from the screen. Then, they were asked to rate the similarity between the next
animation (e.g., SAME MANNER variant) to the Original animation. A 7-point
scale was used, 1 being least similar and 7 being most similar. Then, the participant
watched the same Original baseline animation again, remembering it, then, press
the button to play the second variant animation (e.g., SAME PATH variant) and
rate its similarity to the Original. The order of each variant presented after the
Original baseline stimuli was counter-balanced.
After the similarity rating task in the (non-linguistic) 0L block, participants were
shown and asked to verbally describe each of the Original animations. Korean L2
learners verbally described the motion events in Korean during the 1L block and
in English during the 2L block by switching from Korean to English. However,
English L1 speakers used English in both blocks without switching language, so they
repeated the same session (Figure 1). All utterances by participants were automatically
recorded by a computer and transcribed later for analyses. To minimize the language
effect on perception found in previous studies (Gennari et al., 2002; Papafragou &
Selimis, 2010), where participants linguistically encode the stimuli before they
respond to a non-linguistic task, (non-linguistic) 0L block preceded the 1L and the
2L block in the present experiments. The experiment took approximately 30 min
per student.
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 469
4.4. Coding
Each participant produced 16 utterances that depicted the motion events in the
animations. For each utterance, the main verb was coded as MANNER verb, PATH
verb or Other verb. The example sentence structures of verbal description in Korean
(Table 1) and English (Table 2) are provided. I counted the total number of
MANNER verbs and total number of PATH verbs for the utterances of each
participant and calculated the average scores to determine his or her lexicalization
patterns of motion events.
Table 1. Example sentences in Korean (as described by participants)Sentence structure Example sentence
Path Verbs (PVs)
PV only 떨어지고 있습니다.
PV + Manner modifier 낙하산 타고 떨어지고 있습니다.
PV + Path modifier 썰매를 탄 남자가 나무를 향해 떨어지고 있습니다.
PV+Manner modifier +Path modifier 한 여자가 비행기 위에 메달려 터널 안으로 진입하고 있다.
PV+Path modifier +Manner Modifier 여자가 하늘에서 나무쪽으로 풍선을 메고 떨어지고 있다.
Manner Verbs (MVs)
MV only 한 남성이 이동하고 있습니다.
MV + Manner modifier 사람이 요트를 타고 이동하고 있어요.
MV + Path modifier 나무 옆으로 착륙하고 있습니다.
MV+Manner modifier +Path modifier 남자가 요트를 타고 결승점을 향해 이동하였습니다.
MV+Path modifier +Manner Modifier 어떤 남자가 왼쪽 건물쪽으로 부스터를 달고 날아가고 있어요.
Other Verbs 테러범이다.
Table 2. Example sentences in English (as described by participants)Sentence structure Example sentence
Path Verbs (PVs)
PV only The man enter the house.
PV + Manner modifier A boy passes the cave with his skateboard.
PV + Path modifier A car is going to the tree.
PV + Manner modifier + Path modifier A boy with the rocket is going up to a building.
PV + Path modifier + Manner Modifier A man is falling from airplane with a. parachute.
Manner Verbs (MVs)
MV only A man is moving.
MV + Manner modifier She is landing with her parachute.
MV + Path modifier A ball rolled toward the tree.
MV + Manner modifier + Path modifier A gramma is flying with an airplane to the tunnel.
MV + Path modifier + Manner Modifier A girl is sliding to the bench in a car.
Other Verbs There is house.
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon470
5. Results
5.1. Linguistic encoding (Verbal description task)
In the 1L block, L1 English speakers produced more MANNER verbs (Mean =
0.76, SD = 0.23) than PATH verbs (Mean = 0.17, SD = 0.22), whereas all Korean
L2 learners produced more PATH verbs (Mean = 0.76, SD = 0.19) than MANNER
verbs (Mean = 0.23, SD = 0.19). A two-way ANOVA was conducted with Language
Group (Korean vs. English) and Verb (PATH vs. MANNER) as independent
variables, and mean score of produced verbs as a dependent variable. The main effect
of Language Group was not significant (F(1.30) = 0.139, p>0.05); i.e., the difference
in the mean score of verbs between the two language groups was not significant.
The main effect of Verb was significant (F(1.30) = 34.596, p<0.05. This suggests that
the difference between the production of PATH verb and MANNER verb was
statistically significant. The interaction between Language Group and Verb was
significant (F(1.30) = 155.076, p<0.05). The result suggests that Korean speakers
preferred PATH verbs, whereas English speakers preferred MANNER verbs. This
result confirms the lexicalization patterns of motion events predicted from the
language-specific patterns in that Korean is a PATH-oriented language and English
is a MANNER-oriented language.
In the 2L block, the preference of the L1 English speakers was unchanged (Mean
= 0.73, SD = 0.27 for MANNER verbs; Mean = 0.24, SD = 0.28 for PATH verbs),
but the Korean L2 learners produced more MANNER verbs (Mean = 0.76, SD =
0.20) than PATH verbs (Mean = 0.21, SD = 0.18), in contrary to the 1L block. A
two-way ANOVA was conducted with Language (Korean vs. English) and Verb
(PATH vs. MANNER) as independent variables, and mean score of produced verbs
as a dependent variable. The main effect of Language was significant (F(1,31) =
15.49, p<0.05); the difference in the mean score of verbs between the two language
groups was significant. The main effect of Verb was significant (F(1.31) = 228.798,
p<0.05. This suggests that the mean difference between the production of PATH verb
and MANNER verb was statistically significant. The interaction between Language
and Verb was significant (F(1.31) = 35.160, p<0.05). This suggests that both Korean
and English speakers produced more MANNER verbs than PATH verbs in the 2L
block, but the tendency is stronger for English speakers. These results suggest that
Korean learners could switch to the L2 (English) type of encoding the motion events by
using more MANNER verbs than PATH verbs as English L1 speakers do (Table 3).
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 471
Table 3. Average proportion of MANNER verbs and PATH verbs in each language
block (SD = Standard Deviation)Group MANNER (SD) PATH (SD)
1L
English L1 English 0.76 (0.23) 0.17 (0.22)Korean Korean L2 0.23 (0.19) 0.76 (0.19)
2L
English L1 English 0.73 (0.27) 0.24 (0.28)English Korean L2 0.76 (0.20) 0.21 (0.18)
5.2. Non-linguistic perception
With the Original animation as the basis for comparison, both groups of speakers
rated SAME PATH variants to be more similar than SAME MANNER variants
(Table 4), but the tendency was stronger for Korean L2 learners than for L1 English
speakers.
Table 4. Mean scores of similarity ratings by L1 Korean L2 English learners and
L1 English speakers
Variant
Language group
L1 Korean L2 English (n = 20)
L1 English speakers (n = 13)
SAME MANNER 4.05 (1.00) 3.87 (1.02)
SAME PATH 4.60 (0.75) 4.07 (1.11)
A two-way analysis of variance was conducted with Language Group (Korean
vs. English) and Variant (SAME MANNER vs. SAME PATH) as independent
variables, and similarity rating score as a dependent variable. The main effect of
Language Group was not significant (F(1.31) = 1.237, p>0.05); i.e., the mean difference
in similarity scores between L1 English speakers and Korean L2 learners was not
significant. Both L1 and L2 groups showed similar behavior in the perception of
motion events. The main effect of Variant was significant; SAME PATH variants
were perceived more similar to the baseline Original stimuli with higher similarity
scores than SAME MANNER variants were and the mean difference was significant
(F(1.31) = 9.393, p<0.05). The interaction between Language Group and Variant
was not significant (F(1.31) = 2.014, p>0.05). The results suggest that both L1
English speakers and Korean L2 learners focused more on PATH of motion than
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon472
on MANNER of motion in the non-linguistic 0L condition.
6. Discussion
The first research question asked whether native English speakers and Korean
learners of English differed in their verbal description of motion events in the 1L
block.
Research hypothesis 1 was supported. Results of verbal description in the 1L block
showed language specific differences in lexicalization patterns between English and
Korean. L1 English speakers produced more MANNER verbs than PATH verbs
in their L1 English when they described motion events. L1 Korean speakers
produced more PATH verbs than MANNER verbs in their L1 Korean. The results
confirmed Talmy’s (1985, 2000) typological differences instantiated in Slobin’s (1996)
thinking for speaking in encoding motion events that explain that language specific
treatment of lexicalization differs in English (MANNER-oriented language) and
Korean (PATH-oriented language).
Language-specific lexicalization patterns of motion events in their native languages
in the 1L block conform to the Sapir-Whorf hypothesis: “…users of markedly
different grammars are pointed by their grammars towards different types of
observations and different evaluations of externally similar acts of observation, and
hence are not equivalent as observers but must arrive at somewhat different views
of the world. (Whorf, [1940] 2012, pp 282-283)”. This paragraph, conceivably one
of the most cited in the history of linguistics, is the basis of the argument about
linguistic determinism derived from “the seductive force of habit” (Leavitt, 2011,
p. 146).
The second research question was whether Korean learners of English could
switch to their L2 (English) type when they were asked to describe motion events
in English in the 2L block. Korean learners used more MANNER verbs than PATH
verbs in their L2 (English), whereas they used more PATH verbs than MANNER
verbs in their L1. This difference suggests a switch to the patterns of L2 thinking
for speaking. Based on this result, research hypothesis 2 was supported.
The shift when using L2 suggests that the bilingual processing of the L2 learners
is like that of the two monolinguals (Korean-like in 1L block and English-like in
2L block). This separate processing for each language is in line with the theory of
monolingual mode (Grosjean, 2001).
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 473
Although it was predicted in the research hypothesis that Korean L2 learners
might engage in a direct translation, transferring their native Korean PATH-oriented
descriptions into English, or might exhibit some variety of mixed behavior, the
results of verbal prediction in the 2L block did not show any clear tendency of L1
transfer. This result suggests that the Korean L2 learners in this study showed limited
influence of L1 thinking for speaking patterns when they speak L2. In addition, the
bidirectional transfer observed in L1 Spanish-L2 English bilingual adults (Hohenstein,
2006) was not observed in the present study, although bidirectional transfer of
motion encoding patterns (i.e., L1àL2 and L2àL1) can be expected in bilingual
speakers whose L1 and L2 encode motion differently.
If the structure of a language determines a native speaker's perception and
categorization of experience, i.e., if language affects the way native speakers of the
language perceive the world, will learning an additional language could result in
conceptual changes?
The first possibility is that learning another language does not have any conceptual
effects; new linguistic items—whether lexical labels or grammatical rules—are
learned, and communication with users of another language is achieved, without
any conceptual change. It is possible to learn to communicate in another language
without a change in one’s concepts. In this case, if bilinguals differ from monolinguals,
it is only in how they think about reality in order to encode it for speech, a.k.a.
‘thinking-for-speaking’ (Slobin, 1996), not in how they think.
If this hypothesis is true, the third research question asked whether L2 learners
would shift towards L2-based conceptual representations, and pay more attention
to MANNER of motion in the non-linguistic 0L block. The results showed that they
did not: L1 Korean L2 English learners were more sensitive to PATH of motion
than to MANNER of motion in conceptualization. Although linguistic switch to
L2 did occur in the 2L block, the cognitive restructuring did not occur in the 0L
block for these L2 learners. Learning an additional language did not result in
conceptual changes when language was not used. Interestingly, contrary to the
predictions, L1 English speakers did not pay more attention to MANNER of motion
than to PATH; rather, they paid more attention to PATH than to MANNER as
Korean L2 learners did. I will return to this point below.
Linguistic determinism, the strong version of the Sapir-Whorf hypothesis, holds
that differences of language will be associated with differences of thought; i.e., that
language and thought covary. Determinism goes beyond this assertion to require
that the prior existence of some language pattern is either necessary or sufficient
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon474
to produce some thought pattern. However, results of the non-linguistic condition
of the present study do not confirm the strong version of Sapir-Whorf hypothesis:
Language-specific difference between Korean and English was not observed when
language was not involved in the conceptualization.
To explain the similar behavior patterns in L1 and L2 that focused more on the
PATH of motion in the 0L block, two accounts can be provided. First, PATH can
be considered to be the “core-schema” of a motion event, as Talmy (2000) asserted.
As mentioned in the previous section, there are many L1 studies that show that
children from various L1 backgrounds produced PATH verbs more often than
MANNER verbs (e.g., Goldin-Meadow & Zheng, 1998; Choi & Bowerman, 1991).
Therefore, PATH of motion may be universal property in encoding motion events.
Alternatively, the English speakers demonstrated the reverse-Whorfian effect: they
preferred a periphery encoding strategy in English, i.e., PATH of motion, as was
observed by English speakers in Papafragou et al. (2008) and Trueswell and
Papafragou (2010). It is plausible that the length of stay in Korea of the English
speakers in the present study and their subsequent exposure to L2 (Korean) input,
although they are using English only in Korea, might have affected their conceptual
restructuring to be more sensitive to PATH of motion. Further research is required
with L1 English speakers with different length of stay in Korea to account for the
reverse-Whorfian effect by them.
7. Conclusion
This study attempted to fill a gap in the current literature by examining both
lexical and syntactic representation with regard to description and perception of
motion events in the acquisition and processing of L2 learners in an EFL context.
The approach probed whether conceptual restructuring occurs in a non-linguistic
condition during linguistic switch to L2-type description. L2 learners did now show
cognitive restructuring when language was not involved, even though language
specific lexicalization patterns were observed in linguistic conditions and switch to
L2 canonical type of description (2L block). Cognitive restructuring failed to occur
by L2 learners when language was not involved. This study suggests that linguistic
switch did not evoke subsequent cognitive change in L2 acquisition and processing.
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 475
References
Allen, S., Özyürek, A., Kita, S., Brown, A., Furman, R., Ishizuka, T., & Fujii, M. (2007).
Language-specific and universal influences in children’s syntactic packaging of manner and
path: A comparison of English, Japanese, and Turkish. Cognition, 102(1), 16-48.
Athanasopoulos, P., & Bylund, E. (2013). Does grammatical aspect affect motion event cogni-
tion? A cross-linguistic comparison of English and Swedish speakers. Cognitive Science, 37,
286-309.
Aveledo, F., & Athanasopoulos, P. (2016). Second language influence on first language
motion event encoding and categorization in Spanish-speaking children learning L2
English. International Journal of Bilingualism, 20(4), 403-420.
Bassetti, Benedetta, and Vivian Cook. (2011). Relating language and cognition: The second
language user. In V. Cook & B. Bassetti (Eds.), Language and bilingual cognition (pp.
143-190). New York, NY: Psychology Press.
Berman, L., & Slobin, D. I. (1994). Relating events in narrative: A cross-linguistic developmental
study. Hillsdale, NJ: Lawrence Erlbaum Associates.
Brown, A., & Gullberg, M. (2013). L1-L2 convergence in clausal packaging in Japanese and
English. Bilingualism: Language and Cognition, 16(3), 477-494.
Brown, R. W., & Lenneberg, E. H. (1954). A study in language and cognition. The Journal
of Abnormal and Social Psychology, 49(3), p.454.
Bylund, E. (2009). Effects of age of L2 acquisition on L1 event conceptualization patterns.
Bilingualism: Language and Cognition, 12(3), 305-322.
Bylund, E., & Athanasopoulos, P. (2014). Language and thought in a multilingual context:
The case of isiXhosa. Bilingualism: Language and Cognition, 17(2), 431-441.
Cadierno, T., & Ruiz, L. (2006). Motion events in Spanish L2 acquisition. Annual Review
of Cognitive Linguistics 4, 183-216.
Choi, S., & Bowerman, M. (1991). Learning to express motion events in English and Korean:
The influence of language-specific lexicalization patterns. Cognition, 41, 83-121.
Choi, S., & Lantolf, J. P. (2008). Representation and embodiment of meaning in L2
communication: Motion events in the speech and gesture of advanced L2 Korean and L2
English speakers. Studies in Second Language Acquisition, 30(2), 191-224.
Czechowska, N., & Ewert, A. (2011). Perception of motion by Polish-English bilinguals. In
V. Cook & B. Bassetti (Eds.), Language and bilingual cognition (pp. 287-314). New York,
NY: Psychology Press.
Gentner, D. (2003). Why we’re so smart. In D. Gentner & S. Goldin- Meadow (Eds.),
Language in mind (pp. 195–236). Cambridge, MA: The MIT Press.
Goldin-Meadow, S., & Zheng, M. (1998). Thought before language: The expression of motion
events prior to the impact of a conventional language model. Carruthers and Boucher, 26-54.
Grosjean, F. (2001). The bilingual’s language modes. In J. Nicol (Ed.), One mind, two
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon476
languages: Bilingual language processing (pp.1–22). Oxford: Blackwell.
Hendriks, H., & Hickmann, M. (2011). Expressing voluntary motion in a second language:
English learners of French. In V. Cook & B. Bassetti (Eds.), Language and bilingual cognition
(pp. 315-339). New York, NY: Psychology Press.
Hohenstein, J. (2006). Is he floating across or crossing afloat? Cross- influence of L1 and
L2 in Spanish-English bilingual adults. Bilingualism: Language and Cognition, 9, 249-261.
Lai, V. T., Rodriguez, G. G., & Narasimhan, B. (2014). Thinking-for-speaking in early and
late bilinguals. Bilingualism: Language and Cognition, 17(1), 139-152.
Leavitt, J. (2011). Linguistic relativities: Language diversity and modern thought. Cambridge,
UK: Cambridge University Press.
Lucy, J. A. (1992). Language diversity and thought: A reformulation of the linguistic relativity
hypothesis. Cambridge, UK: Cambridge University Press.
Maguire, M. J., & Dove, G. O. (2008). Speaking of events: What event language can tell
us about event representations and their development. In T. F. Shipley & J. Zacks (Eds.),
Understanding events: From perception to action (pp. 193-220). New York, NY: Oxford Uni-
versity Press.
Malt, B. C., Sloman, S. A., & Gennari, S. P. (2003). Universality and language specificity
in object naming. Journal of Memory and Language, 49(1), 20-42.
Oh, K.-J. (2003). Manner and Path in motion event descriptions in English and Korean. In
B. Beachley, A. Brown & F. Conlin (Eds.), Proceedings of the 27th Annual Boston University
Conference on Language Development (pp. 580-590). Somerville, MA: Cascadilla Press.
Papafragou, A., & S. Selimis. (2010). Event categorisation and language: A cross-linguistic
study of motion. Language and Cognitive Processes, 25(2), 224-260.
Papafragou, A., Hulbert, J., & Trueswell, J. (2008). Does language guide event perception?
Evidence from eye movements. Cognition, 108(1), 155-184.
Papafragou, A., Massey, C., & Gleitman, L. (2002). Shake, rattle,‘n’roll: The representation
of motion in language and cognition. Cognition, 84(2), 189-219.
Slobin, D. (1996). From “thought and language” to “thinking for speaking”. Rethinking
Linguistic Relativity, 17, 70-96.
Slobin, D. (2000). Verbalized events: A dynamic approach to linguistic relativity and deter-
minism. In S. Niermeier & R. Dirven (Eds.), Evidence for linguistic relativity (pp. 107-138).
Amsterdam: John Benjamins.
Slobin, D. (2004). The many ways to search for a frog. In S. Strömqvist & L. Verhoeven
(Eds.), Relating Events in Narrative. Vol, 2, Typological and contextual perspectives (pp. 219-257).
Mahwah, NJ: Lawrence Erlbaum Associates.
Swoyer, C. (2011). How does language affect thought? In V. Cook & B. Bassetti (Eds.),
Language and bilingual cognition (pp. 3-22). New York, NY: Psychology Press.
Talmy, L. (1985). Lexicalization patterns: Semantic structure in lexical forms. In T. Shopen
(Ed.), Language typology and lexical description: vol.3. Grammatical categories and the lexicon
(pp. 57-149). Cambridge: Cambridge University Press.
Language Research 55-3 (2019) 457-477 / Soo-Ok Kweon 477
Talmy, L. (2000). Toward a cognitive semantics: vol. II: Typology and process in concept structuring.
Cambridge, MA: MIT Press.
Trueswell, J, & A. Papafragou (2010). Perceiving and remembering events cross-linguistically:
Evidence from dual-task paradigms. Journal of Memory and Language, 63, 64-82.
Whorf, Benjamin (2012). Language, thought and reality: Selected writings of Benjamin Lee Whorf
(2nd ed). Eds by J. B. Carroll, S. C Levinson & P. Lee. Cambridge, MA: MIT Press
(Original work published 1940).
Soo-Ok Kweon
Professor
Division of Humanities & Social Sciences,
POSTECH
77 Chengam-ro, Nam-gu Pohang, Kyungpook 37673, Korea
E-mail: [email protected]
Received: October 12, 2019
Revised version received: November 7, 2019
Accepted: December 16, 2019